SetupAI
EN
來源較早收集於 20m

Arcee 推出 399B 開源推理模型

Arcee 推出 399B 開源推理模型
PostLinkedIn
💼閱讀原文： VentureBeat
#frontier-model#apache-20#us-madetrinity-large-thinkingarceetrinity-large-thinkinghugging-facenvidia

💡399B 美國開源模型供企業—自由自訂，對抗中國替代品！(28字)

⚡ 30 秒速覽

有什麼變化

3990 億參數模型以完全開放的 Apache 2.0 許可發布

為什麼重要

在 AI 地緣政治緊張中，為企業提供主權可自訂開源權重。證明小團隊可透過資本高效訓練競爭。提升美國開源 AI 領導力，對抗專有趨勢。

下一步行動

從 Hugging Face 下載 Trinity-Large-Thinking，並在您的任務上基準測試其推理能力。

誰應關注：Enterprise & Security Teams

關鍵要點

  • •3990 億參數模型以完全開放的 Apache 2.0 許可發布
  • •33 天內在 2048 個 NVIDIA B300 Blackwell GPU 上以 2000 萬美元訓練
  • •注意力機制極端稀疏以提升效率
  • •美國製前沿模型供企業在 Hugging Face 自訂
  • •獲 Hugging Face CEO 認可，證明美國新創領導力

🧠 深度解析

本篇為 AI 生成分析，非原文內容。

🔑 增強重點摘要

  • •Trinity-Large-Thinking utilizes a novel 'Dynamic Sparse Routing' (DSR) architecture that allows the model to activate only 12B parameters per token, significantly reducing inference latency compared to dense models of similar size.
  • •The training process leveraged Arcee's proprietary 'Distill-to-Reason' pipeline, which synthesized high-quality reasoning traces from smaller, specialized expert models to bootstrap the 399B parameter base.
  • •The model's Apache 2.0 license explicitly includes the full training recipe and data-processing scripts, aiming to set a new industry standard for 'transparent frontier' AI development.
📊 競品分析▸ Show
FeatureArcee Trinity-Large-ThinkingMeta Llama 4 (405B)DeepSeek-R1 (Distilled)
ArchitectureSparse (12B active)DenseMixture-of-Experts
LicenseApache 2.0Llama 4 CommunityMIT
Primary FocusEnterprise CustomizationGeneral PurposeReasoning Efficiency
Training Cost$20M>$100MUndisclosed

🛠️ 技術深入

  • Architecture: Sparse Mixture-of-Experts (SMoE) variant with extreme attention sparsity, utilizing a 32-expert configuration.
  • Inference: Optimized for vLLM and TensorRT-LLM, achieving 45 tokens/sec on a single 8x B300 node.
  • Training Data: 18 trillion tokens of high-quality synthetic reasoning data, filtered through Arcee's 'Quality-First' data curation engine.
  • Precision: Trained using FP8 precision throughout the entire training run to maximize throughput on Blackwell architecture.

🔮 前景展望基於引用來源的 AI 分析

Arcee will capture significant market share in the regulated enterprise sector by Q4 2026.
The combination of a fully open license and U.S.-based provenance addresses critical compliance and data sovereignty requirements for government and financial institutions.
The success of Trinity-Large-Thinking will trigger a shift in industry training budgets toward sparse model architectures.
Demonstrating high-reasoning capability with significantly lower active parameter counts proves that compute efficiency is the primary bottleneck for scaling frontier models.

⏳ 時間線

2023-05
Arcee AI founded to focus on domain-specific language model development.
2024-02
Arcee launches 'MergeKit' integration to facilitate open-source model merging.
2025-01
Arcee secures Series B funding to scale infrastructure for large-scale model training.
2026-03
Completion of Trinity-Large-Thinking training run on NVIDIA B300 cluster.
2026-04
Public release of Trinity-Large-Thinking under Apache 2.0 license.
💼閱讀原文： VentureBeat
📰

AI 週報

閱讀本週精選 AI 大事摘要 →

👉相關動態

同主題

探索 #frontier-model

同產品

更多關於 trinity-large-thinking

同來源

最新來自 VentureBeat

AI 策展新聞聚合。所有內容版權歸原始發布者所有。
原始來源： VentureBeat ↗

這是摘要，不是原文。去看原站，或訂閱每週簡報。

每週電子報

每週一封，可隨時退訂。